Raman spectroscopy based pathological analysis and discrimination of formalin fixed paraffin embedded breast cancer tissue

Abstract The application of Raman microspectroscopy in the histopathological analysis have been widely pursued for characterizing ex vivo biological tissue, and for reducing the number of false positive biopsies and increase the cancer diagnosis accuracy. Many studies have been carried out on fresh or frozen tissue samples, which preserved comprehensive qualitative and quantitative compositional information. However, by the wider applicability of dewaxed formalin-fixed paraffin-embedded (FFPE) samples in the hospital, an abundant tissue bank is a useful resource for the on-site retrospective research to improve diagnosis accuracy. Therefore, we implemented an ex vivo Raman spectroscopic study combined multivariate analysis methods for presenting a better understanding of the sample processing effects and testifying the potential Raman based pathological analysis capabilities of FFPE sections. By comparing 600 spectra from thirty-six fresh frozen (n=18) and FFPE breast tissue samples (n=18), including Healthy tissue, Ductal Carcinoma In Situ, and Invasive Ductal Carcinoma tissue, the results demonstrated that the dewaxing process significantly altered the biochemical composition of the tissues, particularly lipids, proteins, and carotenoids. Even though, the analytical result proved that commonly used multivariate analysis methods, including principal component analysis - linear discriminant analysis (PCA-LDA) and partial least squares- discriminant analysis (PLS-DA), could still distinguish the investigated tissue types effectively with satisfying overall accuracy in which PCA-LDA is 88.3 %, PLS-DA is 93.0 %. Therefore, this study confirmed that FFPE sections have diagnostic potential with multivariate analytical model provided that the biochemical changes, meanwhile tissue processing should be aware.